This invention relates to optical power control arrangements for optical transmission systems, to receivers, optical interfaces, demultiplexers, add drop multiplexers, optical switches, all incorporating power control arrangements, to distributed power control arrangements, to optical power controller components, to methods of controlling optical power, to methods of coupling wavelengths via an optical interface, and methods of offering a data transmission service over optical transmission systems.
It is known to control optical power level in optical WDM (Wavelength Division Multiplexed) transmission systems, and it is very important to control it accurately and consistently. Optical power is a principal factor limiting reach in amplified systems and therefore determining how many repeaters or regenerators are needed. This affects capital cost and maintenance costs. Optical power level also limits data rate and therefore capacity. Optical power can be considered in terms of total optical power of all WDM channels in the same optical path, or power per channel. If total power is too high, effects such as Stimulated Raman Scattering can lead to distortion of channel powers (see copending US patent application entitled xe2x80x9cCONTROL OF SRS INDUCED TRANSIENTSxe2x80x9d Nortel Networks ref 15240). If per channel powers are too high non-linear effects such as Self Phase Modulation (SPM) can limit the achievable bit error rate, non-linear effects such as SPM cannot be compensated for and hence must be avoided. If power per channel is too low, then at the channel receiver at the end of the optical path, the signal will be harder to distinguish from noise, and unwanted bit errors will occur.
Conventionally, power variations were known to include long term variations in total power level, and short term or transient variations. The longer term variations were compensated by amplifier control loops, both feed forward and feed back loops, based on total optical power detected before and after the amplifier, or by adjustment of the transmitter power via feedback control systems. However, the gain of optical amplifiers can only be controlled for all channels, not separately for individual channels. The speed of amplifier gain control is also limited by the characteristics of the control loop and the optical behaviour of the amplifier, similarly changing the power of the transmitter is a slow process due to the long feedback path to instruct the laser to change its power.
So short term variations or transients, are usually not compensated, but are tolerated by providing margins in the power budget at the design stage, and/or by providing for error correction of bits at the receiver for example. U.S. Pat. No. 6,222,652 (Roberts) shows how sources of transients can be detected, and discusses their causes.
Erbium doped fibre amplifiers can cause amplitude transients when amplifying several wavelengths at once. Consider the simple example of two wavelengths, if one wavelength is instantaneously removed while the pump power remains constant, then the output power at the other wavelength will increase by approximately 3 dB. The speed of the output transient is determined by the pump power and by the response of the Erbium doped fibre to changes in input power and is measured in microseconds. Addition of a second wavelength causes a similar 3 dB drop in the output power of the first wavelength present. In networks equipped with Optical Add/Drop functionality the number of wavelengths propagating through a given amplifier can change during the lifetime of a network, due to reconfiguration of the network or through failures within the network. In addition amplifiers and other optical elements have a polarisation dependent gain or loss and changes in polarisation thereby cause changes in power level. Even vibrations and mechanical movement of optical fibres, connectors and other components of the optical path can alter the loss and so cause variations in the optical power.
Further problems concerning optical power level control on a network wide basis are set out in U.S. Pat. No. 6,392,769. Here, it sets out that in each amplifier node of the WDM system it is desirable to maintain the transmitted power within a small range (typically about 1 dB variation). To accomplish power management at a node, it is possible to use an automatic level controller consisting of a power tap into a detector and an adjustable loss element (optical attenuator) with a simple feedback control loop which continuously compares the power level to a desired value, and adjusts the attenuation to maintain the power level at the desired value. However, real WDM networks include cascades of (typically nonlinear) fiber, amplifiers, and loss elements. The resulting system is nonlinear and potentially chaotic, such that when the simple control algorithm described is independently utilized in each span, the result can generate oscillations in power levels which are much worse (larger in amplitude and faster in frequency) than the original level fluctuations which the system is intended to correct. Further, while this kind of problem can occur in single wavelength networks, it is compounded by inter-wavelength interactions in WDM networks, and grows increasingly difficult to predict as the number of wavelengths increases. Finally, networks which deliberately add and drop wavelengths from a WDM span will need to be dynamically reconfigured. This patent proposes addressing the problem by eliminating feedback instabilities arising from interactions between automatic control loops at the amplifier nodes of an optical network by restricting operation to a single amplifier node at a time. Amplifier node activation is accomplished using a global control signal which is sequentially passed from an upstream node through all of the nodes of the system. This can be a very slow process as only one node can change its behaviour at a time.
U.S. Pat. No. 6,304,347 shows some known optical power management strategies for managing signal power levels in an optical network. In one power management strategy, a consistent output power per wavelength is maintained between neighboring network elements in an OBSLR (optical bidirectional switched line ring) network. Consistent means that the signal power level between network elements will not change significantly enough, over any switching condition in the network, to affect the ability of the network to carry traffic. This localizes power management within each node since input power levels to the nodes remain constant. As a result, power management for the network becomes a function of each node""s internal component configuration and optical path variations. In this strategy all switching scenarios are folded into a small set of operating modes.
In another power management strategy shown in this patent, signal power parameters for different network switching scenarios are tracked. Thus, it is possible to optimize the available signal-to-noise ratio (SNR) in the network at the cost of calculating, storing and exchanging signal power parameters around the optical network. In another power management strategy, signal power parameters for different network switching scenarios are pre-computed and stored. The pre-computed values provide a way for network elements to quickly react to switching events without necessarily having to re-compute parameters as each event occurs.
U.S. Pat. No. 6,268,945 (Roberts) entitled xe2x80x9cStable power control for optical transmission systemsxe2x80x9d, shows an optical transmission system, in which changes in optical power are anticipated and damped by controlling the transmitter output power, or an external damping element. For WDM systems, wavelengths can be added or removed without causing rapid changes in total power which would otherwise disturb the output of downstream optical amplifiers, and cause bit errors.
It is an object of the present invention to provide improved apparatus and methods. According to a first aspect of the present invention, there is provided an optical power control arrangement for an optical transmission system, the arrangement comprising:
an optical delay element in an optical path of the system,
an optical power detector upstream of the delay element,
a power adjusting element for adjusting a per channel optical power in the optical path downstream of the delay element, and
a controller for controlling the power adjusting element to maintain the per channel optical power constant, using an output fed forward from the detector.
An advantage of a feed forward arrangement around a delay is that power transients can be detected in time for pre-emptive control to suppress transients which are too fast to be suppressed by conventional feed back control of optical amplifiers. Suppressing such transients is a radical departure from the conventional approaches described above, of providing power margins and/or error correction, and trying to detect transient sources. Notable consequences include the ability to reduce power margins, prevent damage to components such as receivers, and relax the specifications on other components such as the responsiveness of optical amplifier control loops. In WDM networks, limitations on how many wavelengths can be added or dropped at the same time can be relaxed. Similarly in wavelength routed networks, limitations on dynamic adding and dropping of wavelengths can be relaxed. As propagation of transients can be reduced or prevented, more complex networks can be contemplated, with less risk of chaotic instability. Particularly significant is the ability to prevent propagation of such transients between optical networks of different operators. The risk of damage or untraceable errors from such transients has been a major factor in limiting use of optical interfaces between networks. Reducing the need to have electrical interfaces, could thus appreciably reduce costs and improve service.
By ensuring the power falling on the receiver remains constant, and the system can always be operated at the optimum point, regardless of changes in power arriving at the receiving site. Optical powers along a system may evolve very rapidly due to faults, or other mechanisms. Optical amplifiers present in the system may not be able to fully suppress any transients and may inflict cross gain modulation onto other channels which were not originally affected. Clearly if these changes in power propagate to the receiver the effects described above may occur.
As a preferred feature, the power controller comprises a variable optical attenuator. This can be simpler to control and less costly than other possibilities, though if it is to compensate for both positive and negative transients, then the VOA cannot be normally operating at zero attenuation. Therefore there will be a loss of power in the steady state.
As another preferred feature, the power controller comprises an optical amplifier. This enables both positive and negative transients to be suppressed.
As another preferred feature, the control arrangement is arranged to detect an unwanted change in optical power at the detector, determine a compensating change in the control of the power adjusting element, and arranged to synchronise the compensating change with the unwanted change as it appears at the input of the power adjusting element. This can enable transients to be compensated more accurately.
As another preferred feature, the controller has a look up table. This is relatively easy and cost effective to implement.
As another preferred feature, the delay element is a dispersion compensation module. Other alternatives for the delay can include an extra length of fibre or particular waveguide materials for example. This can help keep component count lower, as one module can have the dual function of compensating for dispersion and providing a delay.
Another aspect of the invention provides a receiver having the above control arrangement. This is the part of the system where constant power is most critical, to prevent damage or prevent bit errors. If the power falling on the optical detector in the receiver varies, it is possible to cause damage, or to reduce the performance of the system.
Another aspect of the invention provides an optical interface for coupling one or more third party wavelengths into an optical transmission system, the interface having the above control arrangement for controlling the power of the one or more third party wavelengths. This may be one of the most commercially significant applications of the control arrangement.
Another aspect of the invention provides a wavelength division demultiplexer for demultiplexing a wavelength multiplexed optical signal into demultiplexed signals each carrying a channel or sub bands of more than one channel, the demultiplexer having the above control arrangement for one or more of the demultiplexed signals.
Another aspect of the invention provides an optical add drop multiplexer for adding or dropping optical signals carrying a channel or sub bands of channels to or from a wavelength multiplexed optical signal, the add drop multiplexer having the above control arrangement for one or more of the signals being added or dropped.
Another aspect of the invention provides an optical switch for switching optical signals carrying a channel or a sub band of channels, the switch having the above control arrangement for one or more of the signals being added or dropped.
Another aspect of the invention provides a distributed optical power control arrangement for an optical transmission system, comprising a series of the above control arrangements distributed along an optical path of the system.
This can enable transients to be suppressed and gain to be controlled along the path, so that each control arrangement need perform only a part of the total compensation, potentially easing the requirements of each control arrangement, or enabling more complex networks to be implemented.
As a preferred feature of the distributed arrangement, the system is a wavelength division multiplexed system, the control arrangements being located where the optical signals are demultiplexed for routing. As the signals are typically demultiplexed for routing by optical switches or OADMs, it is beneficial to ensure constant gain at such points to reduce propagation of transients onto different routes. It is harder to suppress such transients when they have propagated onto routes where only some of the wavelengths have the transients. Also, it is preferable to avoid demultiplexing the signal more often than necessary, to avoid performance deterioration and reduce costs. Hence it is beneficial to locate the per channel power control arrangements where the signal is demultiplexed for other reasons.
Another aspect of the invention provides an optical power controller component for use in controlling a power adjusting element located downstream of a delay element in an optical path of an optical transmission system, the component being arranged to use a detection of a power level change upstream of the delay element to control the power adjusting element to compensate for the change when it reaches the power adjusting element.
This aspect is a principal component of the power control arrangement. It can be used as an upgrade to existing systems already having other parts of the arrangement, such as a delay element followed by a power adjusting element.
Another aspect of the invention provides a method of controlling optical power in an optical transmission system, comprising the steps of:
detecting optical power in an optical path of the system upstream of a delay element,
adjusting a per channel optical power in the optical path downstream of the delay element, to maintain the per channel optical power constant, using the detected optical power.
Another aspect of the invention provides a method of coupling one or more third party wavelengths into an optical transmission system via an optical interface, the interface having the above power control arrangement for controlling the power of the one or more third party wavelengths.
Other aspects of the invention include software for implementing the power controller component or for carrying out (which encompasses controlling) the method steps. This acknowledges that such software can be a valuable, separately tradable commodity. A power controller in the form of software is intended to encompass software which runs on or controls xe2x80x9cdumbxe2x80x9d or standard hardware, to carry out the desired functions, (and therefore the software essentially defines the functions of the compensator, and can therefore be termed a compensator, even before it is combined with its standard hardware). For similar reasons, it is also intended to encompass software which xe2x80x9cdescribesxe2x80x9d or defines the configuration of hardware, such as HDL (hardware description language) software, as is used for designing silicon chips, or for configuring universal programmable chips, to carry out desired functions. The controller could be implemented using just analogue and/or digital electronics, with no DSP or software
Another aspect provides a method of offering a data transmission service over an optical transmission system having the above power control arrangement. Another aspect of the invention provides a method of offering a data transmission service using a third party wavelength carried by an optical transmission system and using the above optical interface. As the advantages of the invention can enable a better network, which is more reliable or more flexible, or has greater capacity, or is more cost effective for example, consequently a data transmission service over the network can show a corresponding improvement, and the value of such services can increase. Such increased value over the life of the system, could prove far greater than the sales value of the equipment.